An extension to emotions and personalities model based on multi-agent systems using normal distribution, Fuzzy logic and interval mathematics

This paper consists of an extension to the model of emotions and personalities based on multiagent systems. The initialization of this model has a fixed values, assigned for the personality and the infiuence of the personality on the emotion. This value should vary in the interval [0;1], where the value zero represents almost nonexistence of the given personality or infiuence on the emotions, and one, the total presence. Therefore, when a value is set at the initialization of the model, it will not be able to show precisely the way these emotions and personalities infiuence the decisions of the human being, since each individual does not possess the ability to quantitatively weigh their emotions and personality. For this, the paper presents the initialization using three difierent techniques: Normal Distribution, Fuzzy Logic and Interval Mathematics. Basing in our results, Fuzzy Logic was the more constant technique in all tested scenarios. The purpose of using these techniques is to be able to have a broader and better view of difierent model output results. In addition, it allows a comparison with the results of the original model, in order to try get a better representation of the human model and to have a more robust model that could be applied in other domains.